This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LM3409 Rise Time Very Slow - 400uS!

Other Parts Discussed in Thread: LM3409HV, LM3409

Hi,

I'm using an LM3409HV eval board straight out the box (got 2, seeing same on both.)

I have not changed the values in any way.

The default is set up for 48V in and 42V out.

I have removed Jumper 1 and am feeding in 5V square wave to the EN pin from a sig gen with it's ground referenced to GND on the eval board.

I have 12 white LEDs on the output with fwd voltages of about 3v3 at 350mA.

If I try 12 x 2v2 red LEDs I see exactly the same issue, just a different amplitude.

I have the amplitude fixed by an external voltage (fine control bench PSU) fed into Vadj through a strong resistor divider network - it works fine to control peak pulse amplitude between 0 and 500mA which is our operating range.  (the eval board comes setup for a 1.5A max design)

I'm measuring the current with a new current probe (which I know is good because external parallel Shunt FET mode works very well)

On the EN rising edge the LED V+ voltage shoots up quickly and levels off for about 80us - during this time hardly any current flows through the LEDs

Probing Vgate I see the duty cycle here is quite low.

Then the bigger problem - the LED current ramps up in a time constant shape waveform to reach its maximum but takes 400us!!

The LM3409/HV datasheet (Figures 17 and 19) and HVEVAL board user manual graphs (Figures 6, 7) show a rise time of a 3.5us - why am I seeing 80us delay + 400uS?

Changing the PWM frequency doesn't help.

I've attached a picture showing the ENable pin (RED), LED current (YELLOW) and LED+ Voltage (GREEN) at 1KHz and at 2.5KHz - the rise time doesn't change - it just gets truncated by EN falling edge in the shorter period as PWM frequency is increased.  It needs 400us to allow it to get to it's max.

Going to 20KHz at per data sheet graphs means the current doesn't have time to rise much really,even at vadj = 1.25V(maximum)

Increasing Vadj to max 1.25V helps get rid of the delay 80us where the current doesn't do much but the really slow 400us rise time is still present.

When I switch the setup over to Parallel Shunt FET mode (jumper 1 links pins 2 & 3, feed 5V terminal and move sig gen pwm input onto PWM2 input) it works a treat - the rise time is only a couple of us as per data sheet waveforms.

Unfortunately the product BOM can't tolerate the extra cost of the shunt FET and it's controlling FET.

Help - what's going on?

Thanks